An Education Tsunami
Will on-line courses destroy universities?
Daphne Koller is a famous researcher who is on the faculty at Stanford University. She started her career as a theorist; I especially remember the paper “Constructing small sample spaces satisfying given constraints” with Nimrod Megiddo at STOC 1993. She now has shifted her focus to learning—not learning theory as in boolean functions, but real learning. People learning. A recent paper of hers is “Modeling how students learn to program” and is at SIGCSE 2012; it is joint with Chris Piech, Mehran Sahami, Steve Cooper, and Paulo Blikstein.
Today I want to talk about the potential coming revolution in delivery of knowledge.
Daphne is playing a key role in this. Stanford is one of the centers of the innovative use of on-line social systems for teaching. Instead of only writing papers about learning, an honorable and well respected model, she has decided to actually change education. Essentially she is following following Alan Kay’s brilliant dictum:
The best way to predict the future is to invent it.
Her vehicle for this change is a company called Coursera, which she founded. It is a spin-out from Stanford University, and is officially linked to various other universities already.
Truth in blogging: we just heard the official announcement that Georgia Tech will be part of Coursera:
Today Georgia Tech will announce a partnership with Coursera, the Stanford University online education spinout that has been much in the news lately. We will join a small group of highly respected partner universities, including Stanford, Michigan, Princeton, and Caltech in a bold experiment in the future of higher education. With all the talk about the nature and desirability of change in higher education, I think it is significant that some of the world’s best universities have decided to partner in this way. It also is significant that Georgia Tech is a founding member of this group.
While my institution has made an agreement, I have no idea exactly what it means yet. I thought it was important however to disclose our connection. Update Tue. 5:35pm: we have learned that the New York Times covered this story today.
The former Stanford Professor Sebastian Thrun also has related ideas, and his vehicle is called Udacity, with founding partners David Stevans and Mike Sokolsky. There are others working toward the same, or almost the same, goal: MIT and Harvard have formed edX, a not-for-profit initiative. Well all these efforts all are not-for-profit so far, but edX is planned to stay that way, while Coursera and Udacity hope to eventually We will see.
The Coursera Approach
Here is a nice quotation from Andrew Ng, also of Stanford University and Coursera.
Last year, Stanford University offered three online courses, which anyone in the world could enroll in and take for free. Students were expected to submit homeworks, meet deadlines, and were awarded a “Statement of Accomplishment” only if they met our high grading bar. Offered this way, my machine learning class had over 100,000 enrolled students. To put this number in context, in order to reach an audience of this size, I would have had to teach my normal Stanford class (enrollment of 400) for 250 years.
That is pretty impressive, and is the key driving force behind the revolution. The ability for one person to reach these kinds of numbers seems to be a game-changer.
Ng goes on to make the key point that technology has changed almost all aspects of our economy, but education has basically stayed unchanged. He goes on to say:
I think there is a huge opportunity to use modern internet and AI technology to inexpensively offer a high quality education online. Through such technology, we envision millions of people gaining access to the world-leading education that has so far been available only to a tiny few, and using this education to improve their lives, the lives of their families, and the communities they live in.
Here is a table that summarizes some of the on-line courses that have happened already or are planned for the future.
Why Online Now?
Why the sudden increase in interest in possible online courses? Said differently: why are there multiple new well funded companies and non-profit efforts happening now? We have had video courses for years, so what is the new twist? One way to understand what is happening is to look at several issues that all courses face, offline or online.
How to answer questions from students? If an online class has 100,000 students clearly there are not enough hours in the day for the professor to answer questions. Nor can anyone afford to hire a huge team of teaching assistants to do this. Gone are office hours. So how will this be handled? The key is through social media. All the approaches plan on using discussion system technology to have students help each other with questions. This is similar to the stackoverflow ideas that are used to answer questions in research. The companies claim that the methods work well, but this is certainly one of the new ideas. Thus, online courses are more than “just” videos.
How to grade exams? Again the huge volume of exams will force the use of automation. Clearly multiple choice exams are pretty straightforward, but are they sufficient? I suspect that AI technology will be used more and more to allow non-multiple choice exams. One must be careful with this, but I believe that it should be possible. Currently essays are auto-graded for the SAT, but there are issues with this.
How to avoid cheating? I have talked to the founders of the two companies and they both are well aware of this issue. The best solution I have heard so far is two-fold. One, do not worry about it, since the students are not yet getting full university credit. The other is to have students take the final exam in a special location, with human examiners. This is already done with other mass exams so it seems a reasonable approach.
How to replace the social aspects of teaching? Students seem to get more out of classes than just the lectures and so on. They meet other students, they interact with other people, there is a social educational aspect to standard classes. How will online courses handle this? It is unclear, and I think it is one of the major open questions.
The Muti-Billion Dollar Question
The question is: are these on-line based courses going to replace traditional courses? Will students not need to sit in classrooms in the future, but rather need only a laptop to get a first-rate education? These questions are called the “Tsunami” by the President of Stanford, John Hennessy, himself a former Stanford computer science professor and one of the founders of MIPS Technologies.
Are we as educators going to disappear? Why take my course on combinatorics if you can have Bob Sedgewick, one of the world experts, give the lectures? Indeed.
Here is a cartoon that I made that attempts to present the dilemma facing us at universities. I hope all take it as it is meant—as a light way to raise the key issue. Will universities as we know them today survive? Will some disappear, will some thrive; will all have to change and adapt?
Things Do Disappear
Things do disappear. Gone are buggy whip companies, slide rule companies, and Kodak cameras. Ken remembers how neat it was that Oxford had a Woolworth’s thrift store, since the chain was a mainstay of his childhood in Paramus, NJ, and undergraduate years at Princeton. Yet a wistful verse by W.H. Auden mourning two pubs that used to be on its site ran through his head every time he entered it:
The Clarendon’s gone—I regret her;
The George is now closed and forgot.
Some changes are all for the better—
But Woolworth’s is probably not.
That Woolworth’s closed its doors at the end of 2008, amid the worldwide recession, and the remains of its UK parent group were absorbed by an online retailer. In the US, F.W. Woolworth stayed one of 30 bellwether companies in the Dow Jones Industrial Average until 1991, when it was replaced by Wal-Mart, and it lasted as a chain only six more years. Now one part of it survives as a popular brand you may have on your feet—can you name it?
Most recently gone is the wonderful bookstore chain that was called Borders. It was founded in 1971 by two brothers with that surname, in Ann Arbor where they were attending the University of Michigan. At its peak it employed over 19,000 people throughout the U.S., and more internationally. Counting Waldenbooks stores too, there were over 500 such stores. It is gone now. The last store closed its doors on Sunday, September 18, 2011—see this for details.
There may be many factors that caused it to cease to exist, but certainly the rise of Amazon and e-book readers were major ones. Borders could not, or chose not to, adapt—and it paid the price. It disappeared. I wonder can the same happen to universities? We did raise this question two-and-a-half years ago, in Latin.
Have you used online courses? Will you use them in the future? Are universities in trouble? What do you think?
[added NY Times update]